obscura/vllm
2
0
Fork 0
mirror of https://github.com/vllm-project/vllm.git synced 2026-08-20 20:50:15 +00:00
Code Issues Packages Projects Releases Wiki Activity

[CI] Add weight transfer tests to CI (#39821)

Browse Source 
Signed-off-by: SumanthRH <[email protected]>
Co-authored-by: Robert Shaw <[email protected]>
Co-authored-by: Cyrus Leung <[email protected]>
This commit is contained in:
Sumanth R Hegde
2026-04-16 15:51:45 -04:00
committed by GitHub
co-authored by Robert Shaw Cyrus Leung
parent b16fda62b7
commit adf9bb3c57
2 changed files with 5 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files
.buildkite/test_areas/distributed.yaml
+2
View File
@@ -196,6 +196,8 @@ steps:
- VLLM_ALLOW_INSECURE_SERIALIZATION=1 python3 examples/rl/rlhf_async_new_apis.py
- VLLM_USE_DEEP_GEMM=1 VLLM_LOGGING_LEVEL=DEBUG python3 examples/offline_inference/data_parallel.py --model=Qwen/Qwen1.5-MoE-A2.7B -tp=1 -dp=2 --max-model-len=2048 --all2all-backend=deepep_high_throughput
- pytest -v -s tests/v1/distributed/test_dbo.py
- VLLM_ALLOW_INSECURE_SERIALIZATION=1 pytest -v -s tests/distributed/test_weight_transfer.py
- pytest -v -s tests/distributed/test_packed_tensor.py
- label: Distributed Tests (2 GPUs)(B200)
device: b200
tests/distributed/test_weight_transfer.py
+3
View File
@@ -41,6 +41,7 @@ def create_mock_parallel_config(
config.rank = rank
config.world_size = world_size
config.data_parallel_rank = dp_rank
config.data_parallel_index = dp_rank
return config
@@ -283,6 +284,7 @@ def inference_receive_tensor(
parallel_config.rank = 0
parallel_config.world_size = 1
parallel_config.data_parallel_rank = 0
parallel_config.data_parallel_index = 0
engine = NCCLWeightTransferEngine(config, parallel_config)
@@ -666,6 +668,7 @@ def inference_receive_ipc_tensor(
parallel_config.rank = 0
parallel_config.world_size = 1
parallel_config.data_parallel_rank = 0
parallel_config.data_parallel_index = 0
engine = IPCWeightTransferEngine(config, parallel_config)